Syntheses, kinetics, and mechanism of ligand substitution reactions of

Salt Lake City, Utah84112-1194, and University of Delaware, Newark, Delaware 19716. Received February 25, 1992. Oxidations of monocarbonyl adducts of ...
0 downloads 0 Views 1MB Size
Download PDF
Organometallics 1992,11, 3215-3224

3215

Syntheses, Kinetics, and Mechanism of Ligand Substitution Reactions of 17-Electron Half-Open Chromocene Carbonyl Complexes Jian Kun Shen,+ Jeffrey W. Freeman,* Noel C. Hallinan,t Arnold L. Rhelngold,+*sAtta M. Arlf,* Richard D. Ernst,'.' and Fred Basolo+~t Depertments of Chemistry, Northwestem Unlverstty, Evenston, Illlnois 60208-31 13, University of Utah, Sak Lake CW, Utah 84 112- 1194, and Univmtty of Deleware, Newark, De4ware 19716 Received February 25, 1992

Oxidations of monocarbonyl adducts of half-open chromocenes lead to 17-electron adducts such as Cp(2,4C7H11)Cr(CO)+ and Cp*(8CBHg)Cr(CO)+.X-ray diffraction studies have revealed that in both casea the open dienyl ligands are present in the usual U-shapedconformations. The former, as a CH3CN solvate of the iodide salt, crystallizes in the monoclinic space group C2 with a = 14.593 (3) A, b = 9.390 (2) A, c = 13.300 (2) A, /3 = 114.91 ( 2 ) O , and V = 1652.9 A3 for 2 = 4. Final discrepancy indices of R = 0.064 and R, = 0.051 were obtained for 1197 unique, observed reflections. The latter crystallizes in the orthorhombic space group Atm with a = 21.351 (8) A, b = 8.598 (2) A, c = 9.763 (4) A, and V = 1792.2 A3 for 2 = 4. Final discrepancy indices of R = 0.048and R, = 0.050were obtained for 1074unique, observed data. Unlike related neutral vanadium and chromium complexes, these cations do not undergo CO exchange reactions readily. However, phosphines readily add to the open dienyl ligands, leading to species such as Cp* (l-PMe,Ph-~~-3-C~Hg)Cr(co)+, whose structure was confiied by a diffraction stud of the BF4salt. The compound crystallizes in the orthorhombic space group Pcaal with a = 9.152 (6) b = 19.206 (8) A, c = 15.181 (7) A, and V = 2668.6 A3 for 2 = 4. Final discrepancy indices of R = 0.066 and R, = 0.086were obtained for 1916 unique, observed data. The open dienyl ligands in thew as well as the neutral laelectron complexes are susceptible to lose in the presence of an excess of additional ligands, Through such reactions one may readily isolate [CpCr(CN-t-C4~41+[sF41-.~, [CpCr(CN-t-C4w4]+[CpCr(CO)3]-, [CpCr(DMPE)(CO)2]+[CpCr(CO)3]; and [CpCr(CN-t-C4&)3]2 Structures of the first two complexes have been determined and found to poeses essentially identical cations. For the former, the crystals are triclinic, space group PI,with a = 10.346 (6) A, b = 11.832 (8) A, c = 15.615 (10) A, a = 78.92 (6)O, /3 = 72.03 (4)O, 7 = 88.85 (4)O and V = 1783 A3 for 2 = 2. Final discrepancy indices of R = 0.077 and R, = 0.077 were obtained for 3320 unique, o b ~ e ~ reflections. ed For the latter, the crystals are orthorhombic, space group Atm,with a = 19.785 (4) A, b = 11.837 (2) A, c = 15.515 (3) A, and V = 3633.5 As for 2 = 4. Final diecrepancy indices of R = 0.103 and R, = 0.120 were obtained for 2358 uuique, observed reflections. Kinetic studies of CO substitutionof [(Cp)(3-C6H,)Cr(CO)]+[BF,Iwith an excess of isocyanides or PPh3afford a second-order rate law. This and the activation parameters (AH*= 16.6 f 0.2 kcallmol and A,9 = -22.6 f 0.6 cal/(mol-K)for t-BuNC) suggest an associative mechanism for the CO substitution reaction. The cationic complex reacts more slowly than does ita neutral isoelectronic 17-electron vanadium compound. ESR measurements show that the unpaired electron spin density on Cr is about twice that on V in these two complexes. The retarded rate for CO substitutionof the cationic chromium complex compared with that of the vanadium compound may be rationalized in terms of higher electron spin density and a more crowded coordination sphere for Cr, making it more difficult for the nucleophile to attack the half-filled molecular orbital on Cr than on V.

1,

Introduction Earlier kinetic studies' on 17-electron organometallic complexes have shown that they undergo rapid associative ligand substitution compared to analogous 18-electron complexes. This enhanced associative substitution lability of 17-electroncomplexes was attributed to a more facile nucleophilic attack on the partially occupied metal orbital in 17-electron complexes compared to that on the filled orbital in 18-electron complexes.1d More recently, it has been shown that the 17-electron open and closed vanadocenee react by both associative and dissociative pathways.4a An interesting observation is that (Pdl)2VC0 complexes (Pdl = C a , or substituted pentadienyl ligand) are leas reactive in associative substitutionthan is Cp2VC0, although Pdl ligands generally have greater electronwithdrawing ability than does the Cp ligand. This was rationalized4 in terms of the rigid (Pdl),Vgeometries and reduced spin delocalization making the metal center less susceptible to nucleophilic attack. Our studies on half-open chromocene carbonyl complexes show they undergo CO exchange predominantly by t Northwestern University.

* University of Utah.

8 University of Delaware.

a dissociative mechanism: a normal behavior of 18eleCtron complexes. However, these complexes have an unusual &S)-coordinated Pdl,which is different from qs (U)-Pdl observed in 16-electronchromium and in 17-electron vanadium carbonyl complexes. We found that these 18electron half-open chromocene carbonyls were easily oxidized by NOBF4to the analogous 17-electron complexes. This provides an opportunity for comparing their structures and reactivities with those of analogous 18-electron (1) (a) McCullen,S. B.; Walker, H. W.; Brown, T. J. J. Am. Chem.SOC. 1982,104,4007. (b)Fox, A.; Malito, J.; Po&,A. J. J. Chem. SOC.,Chem. Commun. 1981, 1052. (c) Zielman, P. M.; Amatore, C.; Kochi,J. K.J. Am. Chem. SOC.1984,106,3771. (d) Hershberger, J. W.; Klinger, R. J.; Kochi, J. K. J. Am. C h m . Soc. 1983,106,61. (e) Herrinton,T. R; Brown, T. L.J. Am. Chem. SOC. 1986,107,5700. (f) Shi, Q. 2.;Richmond,T. G.; Trogler, W. C.; Baeolo, F.J. Am. Chem. SOC.1984,106,71. (e) Trogler, W. C., Ed.Organometallic Radical Processes; Elsevier: A". 1990,

and references therein. (2) Harlow, R. L.; McKinney, R. J.; Whitney, J. F. Organometallics

1983, 2, 1839. (3) Narayanan, B. A.; Kochi,J. K. J. Organomet. Chem. 1984,272, c49. (4) Kowaleeki, R. W.; Basolo, F.;Tmgler, W. C.; Gedridge, R. W.; Newbound, T. D.; E m t , R. D.J. Am. Chem. SOC.1987,109,4880. (5) Hallinan, N. C.; Morelli, G.; Basolo, F. J. Am. Chem. SOC.1988, 110.6585. (6) Freeman, J. W.; Hallinan, N. C.; Arif, A. M.; Gedridge, R. W.; Ernst, R. D.; Baaolo, F.J. Am. Chem. SOC.1991,113,6509.

0276-733319212311-3215$03.00/0 Q 1992 American Chemical Society

3216 Organometallics, Vol. 11, No.10, 1992

complexes. The higher oxidation state of the metal atom in these complexes makes them ideal candidah to examine spin localization and steric effects on rates of CO substitution reactions.

Experimental Section General Procedurecl. The half-open chromocenes are very air-sensitive, and sometimes pyrophoric. Their ligand adducts are slightly to somewhat air-sensitive. All compounds were therefore prepared, handled, and stored under nitrogen in a glovebox, while solutions were generally manipulated on a highvacuum or Schlenk lime under Nz, Ar,or CO. Hydrocarbon and ethereal solvents were predried and distilled under Nz from Na/benzophenone prior to use. NMR spectraldata were obtained on a Variau XG300 spectrometer, IR spectral data were obtained on a Perkin-Elmer Model 298 spectrophotometer, and maas spedra were recorded on a VG Micromass 7070 double-focusing maea analyzer with a VG Data System2OOO at an ionizing potential of 17 eV. Magnetic suaceptib~tieawere determined by the Evans method using solution ~ a m p l e s . ~Elemental analyses were obtained from Desert Analytics (Tuceon, AZ)and the Analytische Laboratorien (Gummersbach, Germany). The various dienes, phosphines, and phosphites were either purchased or prepared by published pr0cedw.S Manipulatio~~ of CrC13(THF)9,8dienyl anions,'O and the various half-open chromocenme were performed on a Schlenk line under an atmosphere of p r e p d i e d nitrogen. Unless otherwise stated, the reactions were carried out in 250-mL three-neck round-bottom flasks equipped with magnetic etirring bars and, when nece88(1ly, either pressure-equalizing dropping funnels or solid addition funnels. Solvent transfers were carried out by syringe. When refluxing of the mixture was required, a water-cooled reflux condenser with a nitrogen inlet was attached. Filtrations utilized coarse glasa frits covered with a small amount of Celite. [Cp(2,4-Mepdl)CrCO]+[BF4]-. A 0.50-g (2.1-mol) sample of Cp(2,4Me$dl)Cr(CO) dieeolved in 20 mL of CH&& was cooled to -78 OC in an ice bath. A 0.w(2.1-mol) portion of [NOHBF,] was slowly added to the solution, and the resulting mixture was slowly allowed to come to ambient temperature. The solvent was then removed in vacuo and the residue extractad with 3 X 10 mL of acetone. The acetone solutions often acquire a dark green color. The extra& were fiitered, concentrated to ca.5 mL, and cooled to -80 OC. The product formed orange cryetals which were isolated and dried under vacuum. The product was purified by recrystallization from acetone (yield 25% (0.17 g)). IR data (NyjOl mull): 3115 (ah), 3110 (m), 2022 (e), 1998 (a), 1493 (ah), 1422 (m), 1283 (m), 1117 (ah), 1095 (sh), 1050 (e, br), 1047 (ah), 1033 (sh), 997 (sh), 947 (w), 927 (w), 872 (m), 847 (m), 833 (w) cm-'. FT-IR (acetonitrile): 2018.5 cm-'.Magnetic susceptibility: p = 2.0 pg. Anal. Calcd for ClSH16BCrF40 C, 47.74; H, 4.93. Found C, 47.04, H, 4.89. [Cp(2~cMegdl)CrO]+I-~H~CN. A 1.- (6.7-mol) sample of Cp(2,4M$Pdl)Cr(CO) dieeolved in 50 mL of ether was cooled to -78 OC. A solution of 0.85 g (3.3 m o l ) of Izin 25 mL of ether waeeddeddropwieetothe dution. Thereagenbinatantly"l to form an orangebrown precipitate, while the solution remained yellow. The reaction mixture was then warmed to room temperature and the solvent removed in vacuo. The residue was extracted with 3 X 15 mL of hexane to remove the unreacted starting material, which was recovered in 47% yield by crystallization from hexane at -80 O C . The reaidue was then extracted with 3 X 25 mL of CH3CN, and the extracts were fitered. The filtrate was concentrated to 30 mL and cooled to -30 OC. The product formed orange crystals that were isolated by syringing off the green supematnant. The product could be purified by (7) EV-, D.F. J. Chem. SOC.1969,2003. (8) (a) Jitkm, 0. N.; Bogert, M. T. J. Am. Chem. SOC.1941,63,1979. (b) Jubi, P.; Kohl, F. X. Organomet. Synth. 1986,3,489. (c) Heitach, C. W.; Verkade, J. G. Znorg. Chem. 1962, 1, 451. (d) Frajerman, C.; Meunier, B. Inorg. Synth. 1988,22,133. (9) Collman, J. P.; Kittleman, E. T. Znorg. Synth. 1966,8, 160. Vavoulii, A.; Awtin, T. E.; wee,s. Y . Znorg. Synth. 1960,6,129. (10) (a) Wibon, D. R; Emt, R.D.Orgonomet. Synth. 1986,3, 136. (b)Yasuda, H.; N&, T.;Lee, K.;Nakamura, A. Organometallics 1983, 2, 21.

Shen et al. recrystallization from CH&N (yield 36% (0.52 g) based on recovered starting material; mp 83-85 OC dec). The green supernatant was concentrated to dryneee, yielding a dark green solid that was soluble in THF and was later identified aa (CsHs)Cr12(THF). IR data (Nujol mull): 3095 (w), 3040 (ah), 2245 (w), 2000 (8, br), 1413 (w), 1040 (w), 1017 (m), 980 (w), 947 (m), 873 (w), 855 (m), 843 (m), 835 (w) cm-'. FT-IR(acetonitrile): 2017.5 cm-'. Magnetic susceptibility: p = 2.0 pw Anal. Calcd for C1&19CrIN0 C, 44.13; H, 4.69. Found C, 43.90; H, 4.67. [Cp*(3-MePdl)CrCO]+[BF4r. This compound was prepared by the same route used to prepare [Cp(2,4~Pdl)CrCO]+tBFJ(via the reaction with [NO][BF4]) and was crystallized from acetone/ether (yield 50%; mp 130-140 OC dec). X-ray-quality crystals were grown by gaseous diffusion of ether into an acetone solution of the cation and were orange. IR data (Nujol mull): 3880 (w), 3090 (m), 3035 (ah),2500 (w), 1955 (8, br), 1910 (ah), 1515 (sh), 1492 (m, br), 1402 (w), 1392 (m), 1367 (m), 1283 (m), 1247 (m), 1219 (m), 1163(m, br), 1095 (8, br), 1075 (ah), 1067 (ah), 1045 (8, br), 1025 (ah), 963 (w), 940 (m), 805 (m), 622 (m) cm-'. FT-IR (acetonitrile): 1990.5 cm-'. Anal. Calcd for C17HuBCrF40 C, 53.29; H, 6.31. Found C, 53.65; H, 6.86. [Cp*(l-PcMegh-3-Me-Pdl)CrCOlBF4. A solution of P M e h in 1,2di&loroethane was added to a [Cp*(3-MePdl)CrCOl+[]solution in l,2-dichloroethane. The reaction was followed with an IR spectrophotometer. The absorbance at 1995 cm-' due to the reactant d d as the band at 1907 cm-'increased. When an equal amount of PMepPh had been added and all the reactant converted to the product, the solvent was removed under vacuum. The d u e was washed with pentaue and then with toluene. The red-brown solid left was recryataUiz8d from =/pentane. The structure of this compound was determined by X-ray diffraction (Figure 6). A similar reaction of [Cp(2,4Me2Pdl)CrCO]+[BF4]with PMe2Ph led to a green solid product having vm at 1915 cm-'in 1,2-dichloroethane. Analytical data best fit the mixed-valence but a formulation [Cp(l-PM%Ph-2,4M%Pdl)CrCO]2Cl(BF4)z, definitive assignment must await a single-crystal structural determination. Anal. Calcd for C4zHMBzCrzF80zPzC1: C, 52.22; H, 5.60, P, 6.42; C1,3.67. Found C, 51.96, H,5.56; P, 6.35; C1, 3.49. [C~C~(CN-~-BU)~](THF)BF,. To a solution of [Cp(2,4Mez-Pdl)CrCO]BF4(80 mg) in 10 mL of acetone waa added tBuNC (0.3 mL). The reaction mixture was allowed to stand at room temperature for 30 h, and the cloudy solution was then fiitered. Acetone was removed under vacuum, and the solid residue was recrystallized in acetone/THF/toluene (1,5, and 7 mL, respectively) mixed solvent. The red columnlike crystals obtained were washed with pentane and dried at room temperature under a flow of Nzgas (70 mg, yield 63%). The molecular structure of the product was determined with a Nicolet PI auC, 57.24; H, todiffractometer. Anal. Calcd for C&&BCrF,: 8.06; N, 9.21. Found: C, 57.34, H, 7.98; N, 9.10. IR (acetone): 2156.3 (m), 2098.3 (e), 2057.5 (ah) cm-'. 'H NMR (CD2ClJ: 6 1.49 (8, 36 HI, 1.82 (b, 4 H), 3.68 (b, 4 H), 4.52 (b, 5 H). Reaction of [Cp(2,4-bh#dl)CrCO]+[BF4rwith PhCH,NC. To an acetone solution (10 mL) containing [Cp(2,4-Me2Pdl)CrCO]+[BFJ (100 mg) was added PhCHINC (0.3 mL). The reaction mixture then stood at room temperature for 2 days, during which time the IR band at 2017 cm-' due to CO of the parent compound completely diaappeared and two new bands at 2113 (m) and 2172 (a) cm-' observed. The cloudy solution was fdtered, and acetone was removed under vacuum. The remaining reaidue was a deep red oil, which was washed with THF/toluene (1/33 X 10 mL) solvent. The oily product had a single ESR peak at g = 2.000 in THF solution at room temperature. Attempts were made to red- the compound with magnesium mesh and d u m naphthalenide (Nph). Magnesium meah was added to a THF solution containing the product, and the mixture was stirred at mom temperature for 2 days. The solution dmged color from red to green-yellow. The IR spectrum of this solution showed that part of the product decompoeed and the 1Selectron compound Cp(2,4-Me2Pdl)Cr(CNCHzPh)z6 was not formed. A solution of NaNph in THF was slowly added to an equivalent in solution amount of [Cp(2,4-MezPdl)Cr(CNCH2Ph)z]+[BF4]at -78 OC. An aliquot of the reaction muture ahowed no IR band in the region 2000-2400 cm-'. No further attempt was made to

17-Electron Half-Open Chromocene CO Complexes

Organometallics, Vol. 11, No.10, 1992 3217

Table I. Summary of X-ray Data Collection Parametem for Cp(2,4-C,Hll)Cr(CO)+I-*CHICN (A),[Cp*(3-C,H,)Cr(CO)]+[BF4](B),[Cp*( l-PMe$h-3-C,H9)CrCO]+[BF4](C), CpCr(CN-t-C4H9)4+[BF4]-* THF (D), and CpCr(CN-t -C4Hg),+CpCr(CO),-(E) A B C D E CrCaHSOBF4P CrC17HuOBF4 Cfl16HleNOI C~lsHlsNlOBF4 Cr2C33H4$40S 383.18 521.30 408.22 608.45 650.75 c2 Pnma Pnma Pcall Pi 14.593 (3) 21.351 (8) 9.152 (6) 19.785 (4) 10.346 (6) 9.390 (2) 8.598 (2) 19.206 (8) 11.837 (2) 11.832 (8) 13.300 (2) 9.763 (4) 15.515 (3) 15.181 (3) 15.615 (10) 90 90 90 90 78.92 (6) 114.91 (2) 90 90 90 72.03 (4) 90 90 90 90 88.85 (4) 1653 1792.2 2668.6 3633.5 (12) 1783 (2) 4 ' 4 4 4 Z 2 0.17 X 0.35 X 0.42 0.30 X 0.28 X 0.15 0.08 X 0.38 X 0.48 0.32 X 0.32 X 0.38 0.10 X 0.36 X 0.41 cryst dimens, mm 1.42 1.64 1.11 1.21 1.30 D(calcd), g cm-a 6.60 3.50 5.23 23.9 p(Mo Ka),cm-l 6.28 297 289 297 289 297 temp, K 0.710 73 0.71073 0.71073 0.710 73 0.710 73 A, A 4-45 3-48 4-45 2.5-45 28 scan range, deg 4-48 +h,+k,+l +h,+k,+l +h,+k,+l *h,+k,+l *h,*k,+l data collected 1074 1916 2358 3320 no. of unique obsd data 1197 0.048 0.066 0.103 0.077 0.054 R(F) 0.050 0.051 0.086 0.120 0.077 RJF) 0.76 1.84 0.76 0.74 0.94 NP),e A-3 characterize the products of these reactions with Mg and with NaNph. [(C6H6)Cr(CN-t -BU)~]+[(C,H~)C~(CO),]-. A 0.6-mL (5.3mmol) sample of tert-butyl isocyanide was added to 0.40 g (2.2 "01) of (Cd-15)(CSH7)Crin 20 mL of hexane, and the resulting orange solution was placed into a Fischer-Porter pressure tube under a CO atmosphere. The CO pressure was increased to ca. 5 atm and the tube heated to 70 OC for 2 days (the reaction was initially carried out at 1atm of CO and at room temperature, but it required 1week of reaction time and the yield of product was leas). The orange-yellow product precipitated out of solution during this time. The excess CO was then vented and the product collected on a frit. The product was then washed with 2 X 10 mL of hexane and extracted into 30 mL of acetone. The fiitrate was concentrated to 15 mL and cooled to -80 OC. The orange crystals (mp 162-167 "C dec) were isolated by syringing off the supernatant and dried under vacuum: yield 43% (0.31 8). 'H NMR (acetone-de,ambient temperature): 6 4.68 (e, 5 H, C5H5), 4.31 (8, 5 H, C6H6),1.54 (e, 36 H, C(CH,),). 13C NMR (acetonitrile-d3, ambient temperature 6 246.80 (8, 3 C, CO), 186.78 (8, 4 C, CNR), 88.24 (d of qn, 5 C, C6HS,J = 177,7 Hz), 81.99 (d of qn, 5 C, C5Hs,J = 172, 7 Hz), 58.91 (8, 4 C, C(CH&, 31.06 (q,36 C, C(CH3),, J = 127 Hz). IR (Nujol mull): 3780 (w), 3645 (w), 3620 (w), 3115 (w), 3100 (w), 2150 9), 2095 (8, br), 2045 (sh), 1890 (a), 1770 (8, br), 1367 (ah), 1305 (w), 1260 (w), 1233 (m), 1195 (m, br), 1110 (m), 1040 (w), 1005 (w), 833 (w), 787 (w), 733 (m), 692 (m), 650 (m), 642 (sh) cm-'. FT-IR (acetonitrile): 2155.4, 2006.5, 2063 (sh), 1893.1, 1774.5 cm-'. Anal. Calcd for C93HlBCr2N403: C, 60.91; H, 7.12; N, 8.61. Found C, 60.85; H, 7.01; N, 8.71. [(C,H,)Cr(DMPE)(CO),]+[ (C6Hs)Cr(CO)J. To a 0.40-g (2.2-"01) sample of (C&d(C&)Cr i 10 mL of THF' was added 0.40 mL (2.4 mmol) of DMPE to form a DMPE adduct. The solution was stirred under 1atm of CO for 30 h, resulting in the formation of a red-yellow solution. The yellow product precipitated out of solution after the addition of 20 mL of hexane. The solid was collected on a frit and extracted into 25 mL of acetone. The filtrate was concentrated to ca. 10 mL, and 5 mL of ether was added. The filtrate was then cooled to -30 "C, resulting in the formation of yellow crystals (mp 207-215 OC dec), which were isolated and dried under vacuum: yield 35% (0.20 9). 'H N M R (acetonitrile-d3,ambient temperature): 6 5.01 (t, 5 H, C$H,, J p H = 1.2 Hz), 4.41 (8, 5 H, C5HS),1.81 (apparent t, 6 H, DMPE CH,, JP-H = 5.6 Hz),1.70 (broad s , 2 H, DMPE CH2),1.51 (apparent t, 6 H, DMPE CH3, J p - H = 5.1 Hz), 1.20 (broad s,2 H, DMPE CHJ. '% NMR (acetonitrile-d3,ambient temperature): 6 247.04 (s,3 C, CO), 246.23 (d, 2 C, CO, J p 4 = 60.5 Hz), 90.29 (d of qn, 5 C, CSHS,J = 179.3,7 Hz), 82.76 (d of qn, 5 C, CsH6,J = 171.7, 7 Hz), 29.68 (m, Jpc = 20 Hz), 19.63 (m, DMPE). IR (Nujol mull): 3810 (w), 3630 (w), 3610 (w), 3510 (w), 3480 (w), 3120 (m), 3085

(m), 1950 (a), 1893 (a), 1875 (a), 1855, (sh), 1780 (8, br), 1740 (a, br), 1720 (sh), 1430 (m), 1410 (m), 1307 (m), 1292 (m), 1256 (w), 1228 (w), 1117 (w), 1107 (w), 1087 (m), 1067 (m), 1012 (m), lo00 (m), 950 (ah), 937 (e), 925 (a), 895 (w), 873 (wf, 860 (w), 842 (m), 807 (ah), 797 (m), 742 (m), 690 (m), 648 (8, br), 640 (sh), 605 (m) cm-'. FT-IR (acetonitrile): 1965.5 (m), 1904.0 (sh), 1893.1 (a), 1774.5 (vs) cm-'. Anal. Calcd for C21H28Cr20$2: C, 48.10; H, 5.00. Found: C, 47.97; H, 5.27. [(C,H,)Cr( t-BuNC)&. A 1.0-mL (8.8m"ml) sample of tert-butyl isocyanide was added to 0.50 g (2.7 mmol) of (CsH5)(C5H7)Crin 25 mL of hexane. The orange-red solution that formed was refluxed for 2 days, which resulted in the formation of a deep red solution and a dark precipitate. The precipitate was collected on a frit, washed with 2 X 10 mL of hexane, and extracted into 30 mL of CH3CN. The green fiitrate was concentrated to ca 15 mL, and 10 mL of ether was added. The fltrate was then cooled to -30 OC. The product formed green microcrystals that were isolated and dried under vacuum (yield 3570, 0.35 g; mp 176 OC dec). 'HN M R (acetonitrile-d3, ambient temperature): 6 4.58 (s,5 H, C a d , 1.47 (8, 27 H, t-C,&). '9c NMR (acetonitrile-d3, ambient temperature): 6 183.65 (e, 3 C, CNR), 86.74 (d, 5 C, C5H5, J 176.1 Hz), 57.17 ( ~ , C, 3 C(CH3)3),29.80 (q,9 C, C(CH3),, J = 128.4 Hz). IR (Nujol mull): 3080 (w), 2140 (81, 2085 (81,2040 (a), 1365 (ah), 1269 (m), 1231 (e), 1190 (8, br), 1020 (m), 1011 (m), 863 (w), 842 (sh), 837 (m), 832 (m), 807 (a), 731 ( 8 ) cm-'. Kinetic Studies. Kinetic data were obtained by followingthe decrease of uco bands of the reactants with a Nicolet 5PC-FT-IR spectrophotometer using a 0.2-mm NaCl IR cell. The CO substitution reactions of [Cp(Pdl)CrCO]+[BFJ with RNC and P€& (eqs 4 and 7) were studied in acetone and dichloroethane, respectively, under conditions having concentrationsof phosphorus ligands at least 10 times greater than that of carbonyl complexea. The duociation of CO from [Cp(PMe$h-Pdl)CrCO]+[BF4](eq 10) was examined in THF without any phosphine ligand. Constant temperatures (hO.1 "C) were maintained with a Haake Model F circulator. Solutionsfor ligand substitution studiea were kept in the dark. Plots of In A va time for the disappearance of reactants were linear over 2 half-lives (correlation coefficienta

>0.99). X-ray S t r u c t u r a l Studies. Single crystals of each of the described compounds were obtained by slow cooling of their concentrated solutions in the solvents mentioned in their p r e p arations. Pertinent parameters relating to data collections and refiiementa are provided in Table I. Data for [Cp(2,4-C7Hll)were Cr(CO)]+I-.CH,CN and [Cp*(3-C,H,)Cr(CO)]+[BF,]collected using a Nicolet PI autodiffracbmeter with accompanying software. All calculations employed the Enraf-Nonius SDP programs. Direct methods were used to find the approximate locations of the chromium and iodine (aswell as several lighter)

3218 Organometallics, Vol. 11, No. 10,1992 Table 11. Poiitional Parameterr for the Non-Hydrogen Atoms of ICs*13-CaH.)Cr (CobI+IBFJ Y z atom X 0.25 0.2272 (1) 0.08407 (5) Cr 0.2734 (5) 0.4231 (6) 0.0115 (2) 0.0488 (2)

0.0613 (3) 0.0951 (4) 0.1488 (2) 0.1718 (2) 0.1870 (3) 0.1356 (3) 0.1840 (3) 0.2225 (4) 0.0296 (3) -0.0034 (3) 0.1531 (4) 0.1032 (2) 0.1313 (3) 0.1875 (2)

0.3960 (6) 0.25 0.25 0.3339 (5) 0.3826 (5) 0.25 0.4359 (8) 0.5495 (6) 0.25 0.25 0.25 0.75 0.75 0.75 0.6231 (5)

0.3891 (5) 0.4520 (7) 0.5868 (7) 0.0683 (5) 0.1958 (4) 0.2734 (7) -0.0538 (5) 0.2348 (6) 0.4065 (8) 0.0788 (8) -0,0136 (6) 0.6034 (9) 0.5156 (5) 0.7379 (6) 0.5910 (5)

atom, after which the remaining non-hydrogen a t o m were found readily from difference Fourier maps. In subsequent refinements of positional parameters the function Cw(lFol - IFJ)was minimized. Once the non-hydrogen atom locations had been nearly 6 e d to convergence, difference Fourier m a p were used to locate the hydrogen atom. These a t o m were included in their positions but were not refined. All structural solutions and refiiementa proceeded routinely. Parameters for the latter structure are provided in Tablee-I1 and 111. For [Cp*(l-PMQh-3-MePdl)CrCO]+[BF4]-, photographic data indicated that the crystal powwed mmm h u e symmetry; systematicabsences determined either of the space group PcaP1 (noncentroeymmetric)and Pcam (centrosymmetric). Initially, the absence of mirror-plane symmetry in the propxed molecule and the statistical distribution of E values suggested that the noncentroeymmetric alternative was correct. The subsequent chemically reaeonable and computationally stable refinement in this space group confirmed its choice. The structure was solved by an autointerpreted Patterson projection. Disorder in the BF4- counterion was modeled as a rotational disorder about about the F(2l-B axis. All non-hydrogen a t o m were anisotropicallyrefined, and all hydrogen atom were treated as idealized contributions. An enantiomorphictest b a d on the refinement of a multiplicative term (1.15 (13)) for 4"' indicated that the reportad hand is correct and that the noncentrosymmetzic aseignment was correct. All computations used programs contained in the SHELXTL program library (version 5.1, G. Sheldrick, Nicolet (Siemens), Madison, WI). Atomic coordinates are given in Table IV,and selected bond distances and angles are given in Table V. For [C~C~(CN-~-BU)~]+[BF,]-#THF, a large number of specimens of crystals were screened to determine their suitability for a diffraction study. All diffracted weakly and broadly, most likely the result of a minor lw of THF from the crystal lattice and the dieorder described below. From photographic evidence, mmm h u e symmetry was determined, and systematic absences indicated the space group hBla (noncentroeymmetric)and &mu (centrosymmetric). On the basis of the alignment of the crystallographic and molecular mirror planes, the centrosymmetric altemative was chosen, thereby imposing mirror plane symmetry

Shen et al. on the molecule. In this space group, the methyl groups of one of the crystallographically unique tert-butyl group (C(S) to C(8)) are rotationally disordered about the N(2)-C(4) axis. Additionally, the molecule of THF located during refinement is severely disordered about the crystallographic mirror plane and an adequate model for the disorder could not be constructed. Similarly, the BF4- ion has very high fluorine thermal parameters, indicating that it is also not rigidly fixed in the lattice. No relief of any of these instancea of disorder was found in the noncentroeymmetric space group, and all further computations were to phmu. The methyl group disorder could be resolved into two ideally staggered seta of three methyl groups. The occupancies of the sets were refiied to nearly equal valuea, and they were thereafter each fixed at half-occupancy. No correction for absorption was required, T(max)/T(min)= 1.07. The structure was solved by direct methods. All non-hydrogen a t o m (includingthe half-occupancymethyl carbon atom) were refined with anisotropic thermal parameter, and hydrogen atoms were treated as idealized contributions. Parameters for this structure are provided in Tables VI and VII. A crystal of [CpCr(CN(t-C4H,))4]+[CpCr(CO)1]was mounted on a fine glass fiber with epoxy cement. Both axial photograph and cell reduction programs failed to reveal any crystal symmetry higher than triclinic. The centrosymmetric altemative was initially assum4 this choice was later supported by the chemically reasonable results of refinement. An empirical correction for absorption was applied to the data. The structure was solved by heavy-atom methods. All non-hydmgen a t o m were refined with anisotropic thermal parameters,and hydrogen a t o m were placed in idealiid locations. Several of the methyl group carbon atoms in the t-Bu groups have large thermal parameters, revealing the presence of disorder in their rotational positions. Atomic coordinates and relevant bonding parameters are given in Tables VI11 and IX. All computations used the SHELXTL (version 5.1) library of programs (G.Sheldrick, Nicolet (Siemens), Madison, WI). EPR Spectra. Solutions of the cationic complex in dichloromethane (10-9-10-' M) were prepared in 4-mm quartz tubes fitted with Teflon stopcocks. EPR spectra were recorded at X-band frequency on a Varian E 4 spectrometer at ambient temperature.

Results and Discussion Syntheses and Structures. The monocarbonyl complexes6of half-open chromocenes undergo reversible oneelectron oxidations, leading to 17-electron complexes (eqs 1 and 2). Cp(2,4Me2Pdl)CrC0+ NO+ Cp(2,4-Me2Pdl)CrCO++ NO (1) Cp*(3-MePdl)CrCO + NO+ Cp*(3-MePdl)CrCO++ NO (2) These complexes are notable in that they are rare examples of transition-metal pentadienyl compounds in "higher" oxidation states. They possess one unpaired electron, exhibit typical ESR spectra, and have C-O stretching frequencies more than 100 cm-' higher that those of the neutral species. Due to the paramagnetism

-

-

Table 111. Bond Distances (A) and Angles (de& for [CP*(~-C~H.)C~(CO)~+[BFIICr-CW cr-c(2) Cr-c(l1) cr-c(3) O-c(l1)

1.402 (5) 1.407 (5) 1.506 (6) 1.508 (5) 1.505 (8)

2.195 (4) 2.156 (4) 2.248 (6) 1.858 (7) 1.145 (7) 127.1 (4) 123.7 (5) 118.0 (3) 107.4 (2) 108.5 (3) 108.3 (5)

179.2 (6) 109.2 (6) 112.1 (4) 105.8 (5) 111.5 (7)

17-Electron Half-Open Chromocene CO Complexes Table IV. Atomic Coordinates and Equivalent Isotropic Displacement Coefficients for [Cp*(l-PMegh-3-MePdl)CrCO]+[BF41X

Y

z

U,A2

2019 (2) 3105 (5) 1156 (6) 4747 (19) 4942 (10) 4126 (9) 5036 (9) 1355 (11) 1720 (12) 2377 (8) 1482 (6)

2500 827 (10) 876 (10) 2500 1708 (15) 2500 2500 2600 1561 (14) 1915 (16) 1476 (11) 124 (13) 843 (30) -638 (31) -823 (29) 1467 (11) 60 (13) 796 (16) -567 (14) -787 (15) 325 (33) 541 (29) -1145 (25)

-317 (2) 393 (8) 773 (8) 8516 (25) 8090 (12) 8475 (14) 9205 (11) -1443 (14) -1457 (11) -1568 (8) 376 (9) 1432 (9) 2153 (26) 1051 (24) 1652 (23) 152 (7) 621 (9) 956 (12) -182 (12) 1272 (13) 2164 (22) 1556 (22) 1066 (25)

38 (1) 77 (3) 77 (3) 139 (3) 341 (3) 254 (3) 173 (3) 110 (3) 131 (3) 128 (3) 59 (3) 81 (3) 169 (3) 132 (3) 124 (3) 56 (3) 78 (3) 126 (3) 122 (3) 132 (3) 127 (3) 102 (3) 138 (3)

868 (8)

671 (26) 347 (17) 1480 (17) 2694 (6) 3632 (7) 4217 (8) 3854 (9) 3342 (9) 1299 (19) 109 (14) 890 (23)

Table V. Selected Bond Distances (A) and Angles (deg) for [Cp*( l-PMegh-3-MePdl)CrCO]+[BFJ (a) Bond Distances Cr-CNT 1.860 (10) Cr-C(6) 2.139 (9) C(3)-C(4) 1.38 (2) Cr-C(l) 1.817 (11) C(4)-C(5) 1.35 (2) Cr-C(3) 2.165 (11) C(5)-C(6) 1.41 (1) Cr-C(4) 2.146 (12) C(6)-C(7) 1.54 (1) Cr-C(5) 2.108 (10) CNT-Cr-C(l) C(3)-C(4)4(5)

Bond Angles 111.9 (5) C(4)-C(5)-C(6) 124 (1) C(5)-C(6)4(7)

Organometallics, Vol. 11, No. 10, 1992 3219 Table VII. Selected Bonding Parameters for [CpCr(CN-t-C4H,)1]+[BF4]THF Bond Distances (A) CrC(1) 2.186 (21) CrC(4) 1.938 (13) N(l)-C(10) 1.428 (18) CrC(2) 2.171 (18) CrC(9) 1.951 (13) N(2)-C(4) 1.141 (18) CrC(3) 2.178 (14) N(l)G(9) 1.172 (17) N(2)-C(5) 1.470 (19) Bond Anglee (deg) C(4)Cr-C(4') 77.4 (8) C(9)-N(l)-C(lO) 175.3 (13) C(4)-Cd(9) 77.1 (5) C(4)-N(2)-C(5) 166.4 (14) C(4)-Cd(Y) 124.1 (6) Cd(4)-N(2) 178.7 (11) C(9)-Cr-C(9') 77.7 (8) CrC(S)-N(l) 176.7 (11)

Figure 1. Approximate molecular structure of Cp(2,4-C,Hll)CrCO+.

120 (1) 127 (1)

Table VI. Positional Parameters for the Non-Hydrogen Atoms of [CpCr(CN-t-C4H,)4]+[BF4]-* THF atom X Y z Cr 0.2019 (1) 0.25 -0.0317 (2) 0.0827 (10) 0.0393 (8) 0.3105 (5) 0.0876 (10) 0.0773 (8) 0.1156 (6) 0.25 0.8516 (25) 0.4747 (19) 0.1708 (15) 0.8090 (12) 0.4942 (10) 0.25 0.8475 (14) 0.4126 (9) 0.25 0.9205 i i i j 0.5036 (9) 0.25 -0.1443 (14) 0.1355 (11) 0.1561 (14) -0.1457 (11) 0.1720 (12) 0.1915 iisj -0,1568 (8). 0.2377 (8) 0.1476 (11) 0.1482 (6) 0.0376 (9) 0.0124 (13) 0.1432 (9) 0.0868 (8) 0.0843 (30) 0.2153 (26) 0.0671 (26) -0.0638 (31) 0.1051 (24) 0.0347 (17) -0.0823 (29) 0.1480 (17) 0.1652 (23) 0.1467 (11) 0.2694 (6) 0.0152 (7) O.Oo60 (13) 0.3632 (7) 0.0621 (9) 0.0796 (16) 0.0956 (12) 0.4217 (8) -0,0182 (12) -0.0567 (14) 0.3854 (9) 0.1272 (13) -0,0787 (15) 0.3342 (9) 0.2164 (22) 0.1299 (19) 0.0325 (33) 0.1556 (22) 0.0109 (14) 0.0541 (29) 0.1066 (25) -0.1145 (25) 0.0890 (23) 0.25 0.2842 (23) 0.2655 (29) 0.3571 (34) 0.25 0.3633 (26) 0.25 0.3150 (26) 0.4201 (30) 0.2882 (20) 0.3405 (23) 0.1557 (24)

of these compounds, X-ray diffraction studies were required to establish the mode of pentadienyl bonding. Although mono(liiand) adducts of the half-open chromocenes appear invariably to possess q5-S-pentadienyl coordination,6oxidation to 17-electronmonocations seems

Figure 2. Molecular structure of Cp*(3-C6H9)CrCO+.

to return the open ligand to the more usual U conformation, as demonstrated by structural results on the Cp(2,4-MezPdl)Cr(CO)+and Cp*(3-MePdl)Cr(CO)+salts. Unfortunately, an accurate structural result could not be obtained for the former complex as a result of pseudosymmetry and/or disorder, leading to partially overlapping C& and 2,4MezPdl images. However, the general result (Figure 1and supplementary material) did demonstrate that t15(U)-2,4MezPdlcoordinationwas present, as was an eclipsed conformation (I). In the hope of obtaining more

1

accurate structural data, and of further confirming the

Shen et al.

3220 Organometallics, Vol. 11, No. 10, 1992 Table VIII. Positional Parameters for the Non-Hydrogen Atoms of [CpCr(CN-C-C,H,),It[CpCr(CO),latom X Y z Cr(1) 0.3836 (1) 0.1921 (1) 0.2455 (1) 0.1868 (1) 1.1862 iij 0.7012 (1) 0.1939 (6) 0.3006 (7) 0.2418 (8) 0.2485 (7) 0.2019 (11) 0.1732 (9) 0.1112 (9) 0.2061 (8) 0.2283 (11) 0.1270 (6) 0.1584 (8) 0.3303 (10) 0.1218 (6) 0.2733 (8) 0.3330 (9) 0.2187 (8) 0.7146 (11) 0.3805 (8) 0.1391 (14) 0.8053 (10) 1.3746 (11) 0.0699 (9) 0.7469 (11) 1.3665 (10) 0.1011 (7) 0.6338 (9) 0.3692 (9) 0.1872 (8) 0.6209 (10) 1.3776 (9) 0.3516 (5) 0.0689 (7) 0.3496 (8) 0.4952 (5) -0,0962 (7) 0.2750 (10) 0.4923 (15) -0.1758 (15) 0.3835 (20) 0.4876 (11) -0.1642 (15) 0.1725 (23) 0.5749 (9) -0,0519 (12) 0.2502 (25) 0.3454 (5) 0.2759 (6) 0.3479 (8) 0.4805 (5) 0.3813 (7) 0.2562 (9) 0.5031 (7) 0.3533 (11) 0.0971 (13) 0.5624 (7) 0.3237 (10) 0.2870 (14) 0.4551 (9) 0.5076 (9) 0.2728 (18) 0.2032 (5) 0.3016 (6) 0.5390 (7) 0.1592 (5) 0.4462 (7) 0.7483 (7) 0.2342 (7) 0.5132 (9) 0.7160 (9) 0.0666 (7) 0.5229 (9) 0.7597 (11) 0.1553 (7) 0.8721 (9) 0.3798 (8) 0.2056 (5) 0.0956 (6) 0.5384 (8) -0.0322 (6) 0.1510 (5) 0.7443 (7) 0.1418 (7) 0.0348 (8) 0.8739 (9) 0.2243 (6) -0.1387 (7) 0.7214 (9) 0.0584 (6) -0,0549 (9) 0.7484 (11) 0.2576 (5) 0.7952 (7) 1.0857 (8) 0.2585 (5) 0.5780 (7) 1.0796 (8) 0.1151 (5) 0.7209 (7) 1.0740 (8) 0.4161 (4) -0.0036 (6) 0.3233 (7) 0.4052 (4) 0.3267 (6) 0.3208 (7) 0.1794 (4) 0.3654 (6) 0.6336 (7) 0.1793 (4) 0.0404 (5) 0.6285 (7) 0.8587 (5) 0.3031 (9) 1.0224 (6) 0.3057 (4) 0.4999 (5) 1.0124 (6) 0.7341 (5) 0.0679 (4) 1.0070 (6)

generality of the $(U)-pentadienyl coordination mode in these paramagnetic cations a structural determination was carried out on [Cp*(3-Me-Pdl)Cr(C0)]+[BF4]-, whose greater number of methyl groups was expected to lead to a more ordered structure. This indeed turned out to be the case (Figure 2; Tables I1 and 111),and as for Cp(2,4MezPdl)Cr(CO)+,an eclipsed conformation (I) was ob-

Scheme I. Reaction Mechanism

1

fNR sr-rum

I

-

,CNR~*

+CNR

I,

q '+ ( 18d

served, with crystallographically imposed mirror-plane symmetry. Apparently, eclipsing H-H or CHS-CH3interactions in the two cationic structures are leas severe than the H-CO or CH3-C0 repulsions that would ensue in the staggered forms. The average Cr-C distancesfor the open and closed ligands are similar at 2.190 (3) and 2.214 (3) A, respectively, and can be seen to be slightly longer than the corresponding averages for the 18-electron isocyanide For both complex Cp(rp(S)-C~~)Cr(CN-2,6-(CH~zC~~.s ligands, the longest Cr-C distance involves the carbon atoms (C(3)or C(7)) engaged in the eclipsing interaction. The significance of the eclipsing interactions may also be gauged by the relative tilts of the methyl groups from the ligand planes. As the pentadienyl ligands seem invariably larger than optimal for metal-ligand overlap, significant

Table IX. Selected Bonding Parameters for (CpCr(N-t-C,H,),]+[CpCr(CO)J Bond Distances (A) Cr(2)-C(6) 2.231 (11) 2.169 (9) Cr(2)-C(7) 2.172 (11) 2.164 (9) Cr(2)-C(8) 2.159 (10) 2.185 (13) Cr(2)-C(9) 2.190 (9) 2.194 (11) Cr(2)-C(10) 2.180 (10) 2.196 (9) Cr(2)-C(31) 1.806 (8) 1.931 (7) Cr(2)-C(32) 1.808 (7) 1.943 (8) 1.949 (7) Cr(2)-C(33) 1.830 (9) C(31)-0(31) 1.181 (10) 1.960 (7) 1.146 (12)

1.153 (9) 1.448 (9) 1.166 (11) 1.426 (11) 1.168 (10) 1.456 (10) 1.143 (10) 1.486 (10) 1.153 (9)

Bond Angles (deg)

C(ll)-Cr(l)-C(16) C(ll)-Cr(l)-C(21) C(ll)-Cr(l)-C(26) C(l6)-Cr(l)-C(21) C(16)-Cr(1)-C(26) C(21)-Cr(l)-C(26) C(31)-Cr(2)-C(32) C(31)-Cr(2)-C(33) C(32)-Cr(2)-C(33) Cr(2)-C(32)-0(32)

77.8 (3) 124.2 (4) 78.9 (3) 77.9 (3) 126.7 (4) 76.9 (3) 90.6 (3) 90.5 (4) 89.0 (4)

178.7 (9)

,Ct

Cr(l)-C(ll)-N(1) Cr(l)-C(l6)-N(2) Cr(l)-C(2l)-N(3) Cr(l)-C(26)-N(4) C(1l)-N(l)-C( 12) C(16)-N(2)-C(17) C(21)-N(3)-C(22) C(26)-N(4)-C( 27) Cr(2)-C(31)-0(31) Cr(2)4(33)-0(33)

176.9 (7) 177.1 (7) 178.2 (7) 177.6 (7) 173.3 (10) 166.8 (9) 174.4 (8) 176.6 (8) 178.4 (7) 178.0 (6)

17-Electron Half-Open Chromocene CO Complexes

Organometallics, Vol. 11, No.10, 1992 3221

-0

substituent tilts are generally observed, as indicated in 11,

I1

apparently in an attempt to point the carbon atom p orbitals toward the metal atom." However, in this case the methyl group (i.e., C(4)) lies significantly out of the ligand plane, by 0.16 A, away from the metal atom, corresponding 2100 2000 to tilt of 6.1" in the wrong direction. This may also conIW.0 2200.0 2200 IAVEWYBERS tribute to the lengthening of the C d ( 3 ) bond relative to Figure 3. IR spectral changes for reaction of [Cp(2,4-C7Hll)the other pentadienyl positions. Similarly, C(l0) lies Cr(CO)]+[BFJ with PhCHzNC at 40 OC. further away (0.23 A) from the cyclopentadienyl ligand plane than do C(8) (0.19 A) or C(9) (0.09 A). These obqC30 servations may readily be attributed to a steric interaction between the eclipsed C(4) and C(l0). The fact that C(8) lies nearly a~ far out of the plane as does C(l0) might also be indicative of significant C(S)-CO repulsions. The two ligand planes are tilted from each other by 10.6 (7)O and form angles with the C d ( 1 1 ) vector of 6.1 and 16.7', respectively. As with the isocyanide complex, the twoelectron-donor ligand is therefore seen to be tilted toward the pentadienyl ligand plane. Consistent with other structum,ll the C(l)-C(2)distance of 1.399 (6) A is shorter than the C(2)-C(3) distance of 1.424 (5) A. The presence of the methyl group on C(3) leads to the usual contraction of the C-C-C angle? cf. C(l)-C(2)4(3) = 127.1 (4)", while C(2)-C(3)-C(2') = 123.7 (5)". Substitution Reactions of the 17-Electron ComFigure 4. Molecular structure of [CpCr(NC-t-C4Hg)4]+[CpCrplexes with Weak Bases. The reaction of Cp(2,4(C0)31-. M@dl)CrCO+ with CO in acetonitrile was examined. No CO exchange was observed to occur over a period of 24 h Table X. Second-Order Rate Constants and Activation at room temperature. When the solution was heated to Parameters for Reaction 4 60 "C, the CO peak in the infrared spectrum disappeared T,O C k, M-'8-l LW,kcal/mol A S , cal/(mol.K) and the solution changed color from red to blue. A similar 17.1 2.31 X 16.6 f 0.2 -22.6 0.6 result was observed when the complex was heated under 29.1 7.61 X lod NP Kinetic data showed that the rate of disappearance 40.0 2.03 X lo4 of the CO frequency under an atmosphere of CO (t = 50 "C, kob = 2.03 X lo4 8-l) was identical with that under would be expecteda6Hence, the formulation of this maN2 (t 60 OC, kob 2.08 X lo4 8-l). terial is still unclear. The reactions of Cp(2,4M%Pdl)CrCO+with itmyanides However, when Cp(2,4-Me&U)Cr(CO)+reacts with twere then examined. Since these ligands have strong abBuNC, the product is [CpCr(CN-t-Bu),]+ (eq 4). No other sorbance bands in the C%N stretching region (2000-2400 intermediate was observed during this reaction. It is cm-l), the reactions were conveniently monitored by reCN-t-Bu cording the formation of these bands with an IR speccpCr(cN-t-B~)~+ (4) Cp(2,4-Me2Pdl)Cr(CO)+ trometer. Changes in the IR spectra for a reaction mixture of the complex and PhCH2NCin acetone provide kinetic believed that the rate-determining first step involves nudata which show that the rate of reaction is first order in cleophilic attack of the entering ligand at the metal, which both complex and isocyanide concentrations (eq 3). may generate a 19e transition state (Scheme I). A plot of In k/T vs 1/T affords activation parameters (Table X). d[CdO] The relative low AH* value and the negative AS*value = -kz[CrCO][CNR] (3) dt support the suggested associative mechanism for the rate-determining step. In the proposed mechanism The complex Cp(2,4-Me2Pdl)CrCO+reacts with excess (Scheme I), rapid Pdl slippage occurs following the f i t PhCH2NC, leading to the apparent incorporation of at rate-determining step, allowing for faster subsequent releast 2 equiv of isocyanide (Figure 3).12 Unfortunately, action, in accord with the general observation that more the product ia a paramagnetic oil, and only limited data basic ligands on a metal facilitate ~1ippage.l~The open could be obtained (v(C----N)2113,2172 cm-I; ESR singlet, pentadienyl ligand is finally released from the metal center g = 2.000. While these data would be consistent with a as more t-BuNC enters the coordination sphere. Other formulation such as examples that a pentadienyl ligand coordinated to the no direct evidence for the presence of the 2,4C& ligand metal is more likely to undergo q6 q3 slippage compared could be obtained, and reduction of the product did not l~ related reactions for 18e with Cp are k n o ~ n ,including as proceed readily to Cp(q3-2,4C,H11)Cr-(NCCH2C6H6)2,

-

-

~

(11)(a) E m t , R. D. Struct. Bonding (Berlin) l984,57,1.(b)Emst, R. D. Chem. Rev. 1988,88,1266. (12)Them b aL0 a pollaibility that two producta, perhap isomers, are formed, but thia seem leee likely.

~~~~~

~

~

(13)Freeman, J.; Baeolo, F. Organometallics 1991,10, 266. (14)Pm-Smdoval, M.A.; Powell, P.;DBW, M.G.B.;PeNk, R N. Organometallics 1984,3,1026.Bleeke, J. R; Peng, W.J. Orgotwrnetallics 1986,5,636.

3222 Organometallics, Vol. 11, No.10, 1992

Shen et al.

complexes (see Erperimental Section and eqs 5 and 6). Cp(C6H7)Cr(CO)+ CO (excess) + t-C,HdC (excess) [CpCr(CN-t-C4Hg),] [CpCr(CO)s]- (5)

-

Table XI. ESR Parameters for Phosphine Adducts of CP(~&C,HII)C&O+ ligand P d AC‘lP), Gc PMe2Ph 2.0037 2.008 26 P(n-Bu), 2.0031 2.011 24 2.0038 2.008 26 PCY3 1.9990 1.997 26 PEtS 1.997 33 P(OEth PPh3 1.997 24

+

Cp(s3-C5H7)Cr(dmpe) + CO (excess)

[CpCr(dmpe)(C0)21+[C~Cr(C0)~1(6) Molybdenum d o g s of both products are known, and the nature of [CpCr(CN-t-C,H,)]+[CpCr(CO) 1- has been confirmed by X-ray diffraction (Figure 4).Ii In the absence of CO, the neutral dimer [CPC~(CN-~-C,&)~]~ may be obtained. There is no reaction observed between Cp(2,4-C7Hll)Cr(CO)+and PPh3or phoephite ligands after several hours at room temperature. When the 1/1 ligand/complex “eawere heated in CH&b, they slowly changed color from red to green to blue with concomitant decrease and f d y disappearance of the CO frequency. ESR spectra of the resulting solutions all show doublets. Relevant g values and coupling constanta are in Table XI. These coupling are attributed to the phosphorus atoms of ligands coordinating to the metal atoms, and the similarity to values for the apparent Cp(Pdl)V(PR3) complexes (ca. 26-46 G) might suggest a reaction such as eq 7. However, the oily nature of these products precluded definitive characterization.

-

O g value before heating. *g value after

heating. ‘After heating.

Table XII. Rater of CO Subrtitution and Ligand Additions for Different Complexen complex r, o c &I, 8-1 &2, M-’ s-’ Cp*(3-C,&)CrCOt 38.4 3.0 X lo4 2.0 X lo-‘ (PPh,)‘ cp*(3-C~H~)Cd!o” 37 1.47 X 1p Cp*(1-PMe2Ph-3-C6HB)-40 3.0 X lo4 crco+ (1.5 X 10-6)d Cp(2,447HIJCrCO+ 38.4 3.6 X lo-‘ (PPha 2.78 X lo-‘ (CNCHZPh)’ 2.03 X lo-’ (CN-t-Bu)‘ Cp(2,4C,HlJCrCo” 40 4.04 X lo-’ CP(~,~-C~H~JV(CO)* 40 2.3 X 10-8 1.1 X lo9 (CO)‘ Reference 10. *Reference 8. Nucleophile in parentheses. Rate constant for reaction in 1atm of CO.

Cp(2,4-MezPdl)Cr(CO)++ PRS Cp(2,4-Me2Pdl)CrPR3++ CO (7)

Table XIII. Comparison of Ligand Parameters (A) in CpPdlVCO and Cp*(3-C&f&&O+ CpPdlVCO cp*(3-c~Hg)crco+ M-C(C0) 1.916 (3) 1.858 (7) 2.256 (3) 2.214 (3) M-C(CP) M-C(Pdl) 2.213 (2) 2.190 (3)

Kinetic studies for the apparent substitution reactions (eq 7) were carried out in 1,2-dichloroethaneand under conditions for which the concentrationsof the ligands were at leeet 10 times greater than those of the complexes. The rate constanta were obtained by following the decrease of the bands of vc0 of the reactants and included in Table XII. The reactions of the complexes with PPh3 proceed predominantlyby an B88ociBtive pathway. In contrast, the rate for the reaction of Cp*(3-MePdl)Cr(CO)+with P(OPh), is independent of the ligand concentration (Figure 6). Other reaulta (Table XII) show that the associative rate of CO substitution is faster for Cp(2,4-Me2Pdl)VC0 than for [Cp(2,4-MezPdl)CrCO]+[BF4]-. Since Cr in the cationic complex is more postiviely charged, one might expect that the cationic complex would be more susceptible to nucleophilic attack. From the results of ESR studies (Ab(Cr) = 26 G A b O = 58 G4),spin density on the metal atoms can be estimated. It turns out that the singleelectron density on Cr in ita complex is about 2 times that on the V atom. Since it has been p r o p ~ e e dthat ~ * ~spin localization may retard associative reactions, one might attribute that rate retardation to the greater localization of the single electron on the Cr atom; the higher oxidation state of Cr reducea spin delocalization and destabilizes ita reaction transition state. Although the two complexes have similar coordination environmenta, it is necessary to consider steric effecta on the rate of the reaction. An X-ray structure determination has been obtained for the vanadium complex.*6 The resulta (Table XIII) show that the cationic chromium complex is more crowded than the vanadium complex. This may in part contribute to the reduced reactivity of the chromium complex toward nucleophilic attack.

of the analogous 18-electron chromium complex. The oxidation of the half-open chromocene carbonyl complexea stabilizes the carbonyl group toward substitution. These observations contrast with the fact that Cr(CO)6+l7 is more labile toward substitution than V(CO)elf Cr(CO)B+has only been characterized at -80 “C by ESR spectroscopy, and there is not much information available regarding its

(15) Additional rtructural de* for the isocyanide complexen are presented in the supplementary material. (16) Gadridge, R. W.;Rheingold, A. L.; E m t , R. D. Unpublished resulta.

(17) (a) Bagchi, R. N.;Bond, A. M.; Colton, R.; Lumecombe, D. L.; Moir, J. E. J. Am. Chem.Soc. 1986,108,3362. (b) Pickett, C. J.; Pletcher, D. J. Chem. SOC.,Dalton Tram. 1976,879. (c) Klinger, R. J.; Kochi, J. K. J.Am. Chem. SOC.1980,102,4790.

‘O.OO

1

30.00 a

0 r

X

* 20.00

n 0 Y

10.00

0.00o

~

*

[ L l x 10’ Figure 5. Plot of komvs PR3 concentrations of PPhS(A)and P(OPh)B(m) for CO substitution reactions of Cp*Cr(3-C&)CO+.

The results (Table XII) show that the dissociation rate

of CO from the cationic complex ia about the h e ae that fromthe neutral vandium complex and is slower than that

17-Electron Half-Open Chromocene CO Complexes

Organometallics, Vol. 11, No. 10, 1992 3223

Table XIV. CO Stretching Frequencies of Cp( 1-~-2,4-CTHl1)CflO+ and of Its Phosphine Adducts in

CHIC12 com dex

van. cm-l

~~

'BCO

1972

co

2015

hand VM. cm-' (adduct) P(~-BU)~ 1877 PCY, 1874 P (~ - B U ) ~ 1915 PMezPh 1913 PMePhz 1914 PCYS 1914 PEtS 1911

1111

-

3340

3460

Figure 7. ESR spectrum of Cp*(l-PMezPh-3-C&)CrCO+ in CHzCll at room temperature.

QCIl51

Figure 6. Molecular structure of Cp*(l-PM@h-3-C6Hg)CrCO+.

electronic structure. In contrast, much has been learned about V(CO)@It is well accepted that CO substitution for V(C0)e occurs initially by attack of the entering ligand at the half-filled orbital, which in the metallocene carbonyl complexes is more shielded from biomolecular attack.l* Reactions of the 17-Electron Complexes with Strong Bases. Addition of P(n-Bu), or other phosphine ligands (Table XII) to a solution of [Cp(2,4-Me2Pdl)CrCO]+[BF4]-in dichloromethane or 1,2-dichloroethane leads to an immediate color change from red to green, disappearance of the CO band at 2015 cm-', and appearance of a new band around 1914 cm-l (Table XIV). Similar behavior is seen for Cp*(3-C6Hg)Cr(CO)+.ESR data for these products are included in Table XI. Although such nucleophilicadditions to coordinated r-hydrocarbons are we were surprised to see that small changes in basicities of the nucleophiles result in different reactions. Furthermore, the role which is defined for 1% electron complexes,2othat nucleophilic attack occurs preferentially at the terminal carbon atom of an open r-hydxocarbon (vide infra), still applies to these 17-electron complexes. Notably, when purified red-brown [Cp*(lPM@h-3-C$Ig)Cr(CO)]+[BF4]- was dissolved in CH2C12, a fast revem reaction was observed. It appears in this case that there is a fast, reversible reaction in solution (eq 8). Cp*(3-MePdl)Cr(CO)++ PR3 * Cp*(l-P€&3-MePdl)Cr(CO)+ (8) The weaker binding of PI&by the Cp* complex may easily be rationalized by either steric or electronic arguments. However, the situation actually appears more complex, as the cation colors are quite different, and analytical and ESR data for the green complex suggest an entirely different formulation. Apparently, the presence of the bulky, electron-donating Cp* group prevents more complex re(18)Kowaleaki, R. M.;Baeolo, F.; &borne, J. H.;Trogler, W. C. Orgammetallrca 1988, 7, 1425. (19)Kane-Maguire, L.A. P.;Honig, E. D.; Sweigart,D. A. Chem. Ref. 1984,84,525. (20) Daviee, 9. G.; Green, M.L. H.;Mingoe, D.M. P. Tetrahedron 1978,34,3047.

wan

u s

Figure 8. ESR spectrum of Cp(l-PMezPh-2,4-C7Hl1)CrCO+ in CH2C12at room temperature. Table XV. Rate Constantr and Activation Parameten for CO Dissociation from the P W P h Adduct of Cp(2,4-C$&1)CflO+ in THF Solution

T,OC

kl, 8-1

31.6 40.1 48.8

2.37 X 7.61 X loa 19.0 X loa

AH*,

kcal/mol 23.0

* 1.4

As*,

cal/(mol.K) -4.3 f 4.5

actions from following the initial nucleophilic addition. An X-ray molecular s t " determination was carried out on [Cp*(l-PM@h-3-C&CrCO]+[BF J. The mulb show that PM%Ph is attached to a terminal carbon atom of the 3-MePdl ligand (Figure 6, Tables IV and V). While few reliable comparisons may be made for these structural data, it can be noted that the Cr-CO distance here (1.817 (11)A) seems to be shorter than that in Cp*(3-CeHg)Cr(CO)+ (1.858(7) A). The isolation of the PMe8h adducta raisee yet another question regarding CO substitution with RNC. The weak base ligands might first attack at the terminelcarbon atom of Pdl, and then the CO substitution occurs by aseociation mechanism or by ligand migration from Pdl to the metal. Both of these mechanisms agree with the observed second-order rate law. To test these possible mechanisms, kinetic studies on the loss of CO from the P M e h adducta have been carried out in THF solvent. The ESR spectra of the kinetic products do not show phosphorus spin

3224

Organometallics 1992,11, 3224-3227

coupling (Figures7 and 8),and the rates of disappearance of the carbonyl complexes are first-order in complex concentrations. A plot of In (&/TIVB 1/T permits an estimate of activation parameters (Table XV). The AH*value (23.0 kcal/mol) is compatible with that for dissociation of CO from analogous complexes,4e6and the A S value is near zero (-4.3 cal/(mol-K)). These results are in accord with a dissociative mechanism (eq 9). The kl

Cp*(l-PMe2Ph-3-MePdl)Cr(CO)

40

+

Cp*(l-PMe2Ph-3-MePdl)Cr+ + CO (9) rate of the reaction was retarded by added CO, which is consistent with a dissociative equilibrium (Table XII). Although the final products were not isolated and characterized, they cannot be converted to the parent complex under a CO atmosphere (1a h ) . These results imply that CO substitution does not first take place by RNC attack on the pentadienyl followed by RNC transfer to the metal but rather by direct nucleophilic attack of RNC at the metal center. Since no RNC adduct was observed, K would be